|
jmax -> RE: MMOWGLI and the Maritime Singularity (4/11/2017 2:44:53 PM)
|
Hi kevinkin [8D]
I wonder if any other coauthors of AP 19 are on this forum? DARPA just released an RFI that touches on some of the ideas we discussed there...
https://www.fbo.gov/spg/ODA/DARPA/CMO/DARPA-SN-17-41/listing.html
Video game developers, we want your thoughts on approaches to scalable, interactive gaming/wargaming that can simultaneously span a large number of space and time scales. We are exploring concepts for games that can help assess a wide range of competitive strategies and outcomes, and that bridge strategic, operational, and tactical gaming.
Of particular interest is Focus Area 3 from the information package:
There is considerable development of game AI to make a balanced, challenging, and enjoyable game experience, as well as recognition that competitive gaming provides a rich corpus of data to develop more capable AI in a general sense.ii DARPA is interested in (1) approaches to both learn and drive novel team decision making across all levels of gameplay (strategic, operational, tactical) using novel AI concepts; and (2) ways to abstract and automate human decision making to construct more interesting and challenging scenarios. For example, it may be possible to use human-trained AI’s to inject change into pre-existing scenarios in unexpected ways or even define new scenarios. AI’s could also serve as managers or cognitive aids to players at any level of gameplay.
These new concepts may build on or require fundamental advancements in underlying mathematical representations of context, joint decision making, or game theory. Respondents should discuss their approaches to achieving qualitative improvements in conventional game artificial intelligence/non-player character (AI/NPC) implementations, as well as how such approaches may allow unprecedented exploration of competitive strategic decision making.
|
|
|
|
